The extent and effect of inbreeding in natural populations remain largely undetermined. Pikas Ochotona princeps have been considered a likely candidate for close inbreeding in natural populations due to observations of frequent juvenile philopatry (colonization of natal home range or neighbouring home range) and high levels of spatial overlap and social tolerance between neighbouring individuals of the opposite sex. A 4-year investigation of inbreeding in pikas, however, has revealed that dispersal and mating patterns are uncoupled in this species, i.e. explained by different hypotheses. DNA fingerprinting analysis revealed that band-sharing scores between mated pairs, identified via parentage analysis, were not commensurate with band-sharing among known first-order relatives, but were similar to scores for a small sample of known second-order relatives (i.e. half-siblings grandparent-grandchild pairs). Band-sharing scores between mated pairs were then compared with those between potential mated pairs within the population to assess whether mating was random or nonrandom with respect to genetic similarity. The results of Monte Carlo randomization tests show that pikas mated with individuals with intermediate genetic similarity in greater proportion than would be expected by chance. These data suggest mate choice in pikas may be based upon intermediate levels of relatedness.